NCEP Modeling Synergy Meeting Highlights: February 26, 2007


This meeting was led by Bill Bua and attended by Eric Rogers, Mark Iredell, Dan Pawlick and Ed Danaher. UCAR/COMETís Stephen Jascourt attended by remote conference.


1. CCS


No one from NCO was at this meeting.However, Eric Rogers reported that the 06z NAM on Saturday, 24 February failed at 57 hours and continued to fail from restarts.NDAS seems to have been the problem, since starting from GDAS initial conditions resulted in no failure.It is not known for sure what caused this failure, since so far no one has been able to replicate it.GDAS has been used as a failsafe for initial conditions through today (26 February). Plans are to go back to NDAS, which was still running even though the initial conditions werenít being used by the NAM as soon as possible (note:initial conditions went back to NDAS at 12z Tuesday 27 February 2007).The mesoscale branch will continue to work on finding out what caused the NAM run to fail.


The NCEP web server continues to have problems at specific times (e.g. slow to no service from 10-11 a.m. EST).Lou Cano and Ricardo Romero are trying to find out what the problem is.


Dave Michaud has left his position at NCO; Brent Gordon will be performing his duties in the interim until a permanent replacement is named.




2a. Global Modeling Branch:


Mark Iredell reported that the retrospective runs of the GSI+hybrid coordinate are almost complete. See January 2007 meeting notes for specifics on what periods are covered by these retrospectives. The codes were passed to NCO as of last week. AIRS and COSMIC data are now ready for use and are being incorporated into the more recent runs. Brent Gordon will invite the operational centers to evaluate the parallel.A definitive date for implementation has not yet been set, but it will have to be prior to the start of the 2007 hurricane season (May 15 in east Pacific, June 1 in the Atlantic).


2b. Mesoscale Modeling Branch:


Other than the NAM run failure noted above, Eric Rogers reported that the NAM domain will be expanded in the next implementation, which will take place after the Global GSI+hybrid implementation stated above.Regarding the hi-res window runs, the plan mentioned last meeting (run 2 bigger regional domains 2x per day) is expected to be implemented as well.A glitch has come up however in recompiling the code for the larger 2x per day runs; precipitation has dropped by a factor of about 2.This may be the result of running 3-year-old code, so the first fix will be to update to the current NAM code in the hi-res window runs.

The unified post will be placed in parallel this summer.This will ultimately result in the same outputs (with a few exceptions based on availability of directly forecast variables in the different models) from all NCEP NWP models.


2c. Global Ensemble Prediction System:


The 20 member (+ control) ensemble parallel is now running and is expected to be implemented late in March.


2d. Short Range Ensemble Prediction System:


Nothing reported.


2e. Marine Modeling and Analysis Branch (MMAB):


Nothing was reported at this time.


3. Feedback from operational centers


Nothing was reported at this time.


4. The next meeting will be held Monday, March 26, 2007 at noon in EMC Rm. 209, with remote conference capability.